>>>I don't have a solution with the existing technology. This
>>>sounds like an excellent performance enhancement request to
>>>Oracle. They shouldn't need to write dirty buffers for the
>>>table being truncated, but they will need to be marked
>>>non-dirty or invalid

Is there evidence that the dirty buffers are being written? It
should only shoot down the buffers in each SGA. Depending on
what is going on at the time, that can be quite expensive...
certainly if there are 1 million buffers in 4 RAC instances...
That is a lot of chain walking
--
http://www.freelists.org/webpage/oracle-l